1353
An Electron-Impact Study of Ionization and Dissociation in Methoxy- and Halogen-Substituted Methanes’ R. H. Martin, F. W. Lampe, and R. W. Taft
Contributionf r o m the Whitmore Laboratory, Department of Chemistry, The Pennsylvania State University, University Park, Pennsylvania. Received November 20,1965 Abstract: Appearance potentials and standard heats of formation of a number of methoxy- and halogen-substituted methyl ions have been determined. Combination of the data with known free-radical ionization potentials results in AHr(CH30.) = + 5 f 2 kcal/mole, AHt(CH30CH2.) = - 12 3 kcal/mole, and D [ C H 3 0 C H r H l = 82 f 3 kcal/mole. A similar evaluation of the available data for the fluoro-substituted methyl ion gives AHf( F C H 2 - )= -19 & 3 kcal/mole and D[FCHrH] = 89 f 3 kcal/mole. Stabilization effects in the polymethoxymethyl series display marked saturation. In the polyfluoromethyl series stabilization of CF3+ is apparently less than that for HCFz+. The recently determined upper limit of 9.5 ev for the ionization potential of the CF3 radical is reconciled with a number of appearance potentials of this ion and the CF3-H bond dissociation energy determined from kinetic data. The results obtained for the other halogen compounds appear subject to larger uncertainties.
s part of A determined
a program concerned with substituent stabilizing effects on gas phase ions,2 we have the appearance potentials of a number of substituted methyl ions arising from the dissociative ionization of some methoxy- and halogen-substituted methanes. As no comparable studies have been reported for the methoxy-substituted methanes and as the electron-impact results for the halomethanes are in many cases meager or we report herewith our results. The results of this investigation provide new information on the stabilizing effects of multiple substitution in the methyl ion and on the relationship between substituent effects in corresponding ion and radical. Experimental Section
All appearance potential measurements were carried out in a Bendix Model 14-101 time-of-flight mass spectrometer which had been modified by the incorporation of differential pumping on the source region and by an electron gun modification permitting the use of the retarding-potential-difference technique. l 2 The potential configurations of the electron gun and source region and the details of the measurements were identical with those reported previously. l S The energy spread of the pseudo-monoenergetic electron beam was approximately 0.1 ev, and either xenon, krypton, nitric oxide, or carbon monoxide (depending upon the appearance potential and mass of the ion being examined) were used to calibrate the electron energy scale. (1) Based upon a thesis submitted by R. H. M. to the Department of Chemistry of The Pennsylvania State University in partial fulfillment of the requirements for the degree of Doctor of Philosophy, 1965. (2) R. W. Taft, R. H. Martin, and F. W. Lampe, J. Am. Chem. Soc., 87, 2490 (1965). (3) R. F. Baker and J. F. Tate, Phys. Rev.,53, 683 (1938). (4) J. W. Warren and J. D. Craggs, “Mass Spectrometry,” Institute of Petroleum, London, 1952, p 36. ( 5 ) J. B. Farmer, I. H. S. Henderson, F. P. Lossing, and D. G. H. Marsden, J. Chem. Phys., 24, 348 (1956). (6) H. Gutbier, 2. Naturforsch., A9, 348 (1954). (7) (a) R. I. Reed and W. Snedden, Trans. Faraday Soc., 54, 301 (1958); (b) ibid., 55, 876 (1959). (8) R. E. Fox and R. K. Curran, J . Chem. Phys., 34, 1595 (1961). (9) R. K. Curran, ibid., 34, 2007 (1961). (10) A. G. Harrison and T. W. Shannon, Can. J . Chem., 40, 1730 (1962). (11) D. L.Hobrock and R. W. Kiser. J . Phvs. Chem.. 68. 575 (1964). (12) R. E. Fox, W. M. Hickam, D. J. Gro;e, and T: Kieldaas, Re;. Sci. Instr., 26, 1101 (1955). (13) G. G. Hess, F. W. Lampe, and L. H. Sommer,J. Am. Chem. Soc., 87, 5327 (1965).
Methylal, trimethyl orthoacetate, methyl acetal, and methyl ketal were Eastman Organic Chemicals; trimethyl orthoformate was obtained from Kay-Fries Chemicals, Inc. ; methylene chloride, chloroform, and methyl iodide were Fischer Certified Reagents; and a sample of tetramethyl orthocarbonate was kindly donated by Dr. Brian G. Ramsey. All the above compounds were fractionated on a Nester and Foust 27-in. spinning-band column and collected over type 4A molecular sieve. The compounds were degassed at least three times at about 10-6 torr prior to introduction into the mass spectrometer. Dimethyl ether, methyl chloride, carbon monoxide, nitric oxide, xenon, and krypton were obtained from the Matheson Co., Inc.; methyl bromide was obtained from Dow Chemical Co.; fluoroform was obtained from Peninsular Chemical Research, Inc.; and difluoromethane was a gift from Dr. Alfred C. Haven of E. I. du Pont de Nemours and Co., Inc. All gases were subjected to the same degassing procedure but otherwise were used as received. Methyl ethyl ether was synthesized by the method of Williams011.l~ After fractionation (bp 8-11 ”) and degassing, mass spectral analysis indicated a purity greater than 98%. Methyl nitrite was prepared by the method of Slaterls and fractionated over two icewater traps into a Dry Ice-acetone trap. After a subsequent fractionation at - 127” it was degassed. Mass spectral analysis indicated a purity in excess of 95 %. Methyl fluoride was obtained from Columbia Organic Chemical Co., Inc. Mass spectral analysis indicated the presence of some high molecular weight impurities which were removed by gas chromatographic separation on a high activity silica gel column. The mass spectrum of the resulting methyl fluoride was identical with that reported in the “API Tables of Mass Spectra.”
Results Methoxy-Substituted Methanes. The appearance potentials determined for the series of substituted methoxymethanes and the standard heats of formation of the ionic species are shown in Table I. In the calculation of the ionic heats of formation it is assumed that no excess energy is involved in the simple dissociation RIRz
+ e +R1+ + Rz + 2e
(1)
so that the energetic relationship between appearance potential and thermochemical heats of formation may be written as the equality
In using eq 2 to calculate AHf(R1+), which is in reality (14) A. Williamson, Ann. Chem. Pharm., 81, 77 (1852). (15) W. K. Slater,J. Chem. Soc., 117, 588 (1920).
Martin, Lampe, Tuft
Methoxy- and Halogen-SubstitutedMethanes
1354 Table I. Electron-Impact Energetics in Methoxy-Substituted Methanes
Molecule
Neutral assumed
AP,Q ev
CHIONO CHaOCH3 CHaOCHKH3 (CH30)zCHz (CHsO)CH? (CH30)zCHCHa (CHIO)~CH (CHD),CH (CH~O)JCCH, (CH30)& (CH302)zCHCH3 (CHD)JCCH~ (CHaO)nC(CH3)2
CH3O H CH3 CHsO H CH3 CHjO H CH3 CH30 CH30 CH3O CHiO
11.07 0.06 11.42 i 0.01 11.25 f 0.07 11.41 i 0.03 10.38 i 0.03 10.34 f 0.07 10.36 i 0.06 10.39 i 0.05 10.39 f 0.10 10.32 f 0.10 10.63 f 0.04 10.37 0.02 10.28 =k 0.05
Ion
*
AHf(ion), kcal mole-‘
~~
NO+ CH30CHz’ (CH30)2CH+ (CH30)aC’ CHaOC’HCHs (CH30)zC+CH3 CH30C+(CH3)r a
The indicated uncertainties represent average deviations from the mean of replicate measurements.
an upper limit, it is obviously necessary to know the heat of formation of the neutral fragment, formed in the dissociation (l), as well as the heat of formation of the parent molecule. As far as the radical heats of formation are concerned, the values AHf(H) = 52.1 kcal/mole16 and AHdCH3)17s1*= 33 i 1 kcal/mole are well established. There is more uncertainty concerning AHf(CH30). Graylg has examined the available data on the kinetics of decomposition of methyl nitrite and methyl nitrate and using the method of SzwarcZ0 has concluded that AHf(CH3)0 = -0.5 f 3 kcal/mole. D’Or and Collin?’ have measured the appearance potential of NO+ from methyl nitrite to be 10.7 ev, which upon taking the ionization potential of NO as 9.25 ev’2’23 leads to the value of -3 kcaljmole. We have remeasured A(NO+) from methyl nitrite as shown in Table I, and our value leads to AHf(CH,O) = +5 kcal/mole. It is likely that within the combined experimental errors all values are in agreement. However, in the calculations of the ionic heats of formation in Table I, we have used our value of AHf(CH30) = 5 kcal/mole and we note that a substantially lower value of AHf(CH30)(say AHf(CH30) < 0) would not give as good an agreement for the heats of formation of ions derived from diferent sources. Hence, we believe that AHf(CH30) > 0, and that within the combined experimental errors our value of +5 kcal/mole and Gray’s estimate of -0.5 kcal/mole are in agreement. That the formation of NO+ from C H 3 0 N 0 involves little or no excess energy is indicated by the observations that (1) the ionization efficiency curve of NO+-is simple, that is, there is no long tail characteristic of an excess energy process; and (2) NO+ is the major ion in the mass spectrum, while C H 3 0 is ~ only a very minor constituent. The molecular heats of formation (in kcal/mole) used in constructing Table I are as follows: AH,(NO) = 21.6,24 AHf(CH30NOj = -- 14.9,24 AHf(CH30CH3) H. A. Skinner and G. Pilcher, Quart. Rev. (London), 17, 264 G. C. Fettis and A. F. Trotman-Dickenson, J . Chevr. Soc., 3037
B. E. Knox and H. B. Palmer, Chem. Rev., 61, 247 (1961). P. Gray, Trans. Furaday Soc., 52, 344 (1956). M. Szwarc, Chem. Rev., 47, 76 (1950). L. D’Or and J. Collin, Bull. Soc. RoJ,.Sei. Liege, 22, 285 (1953). E. M. Clarke, Can. J . Phys., 32, 764 (1954). G. G. Cloutier and H. I. Schiff, J . Cheri. P h j s . , 31, 793 (1959). C. T. Mortimer, “Reaction Heats and Bond Strengths,” Addison: Publishing Co., New York, N. Y . , 1962, Chapter 7.
Journal ofthe American Chemical Society
/ 88.7 / April 5, 1966
b
5b
167 173 172 101 113 112 66 74 68 146 131 99
4Hf(CH3O).
= -44.3,24 AHf(CH30CZH5) = -53.2,24 AH,[(CH30)2CH2] = -86.1,’’ AHf[(CH30)2CHCHS] = ~ 9 3 . 1 , ’ ~AHf[(CH30)3CH] = - 121.3,“ AHf[(CH30)3CCH3] = - 133.0,” AHf[(CH30j,C(CH3)2] = - 102.3,27and AHf[(CH30)4C]= - 164.5.?’ rn/e 45. CHBOCHz+. There is no doubt of the identity of this ion from CH30CH3 and little doubt when it arises from (CH30)2CH2. There is a possible ambiguity, however, in the case of m/c 45 from CH30CH2CH3since there is the possibility of forming CH3CH20+. We can rule out this possibility, however, since AHf(C2H60+)is known. The measurements of Kanomata28on (C2HJ20 give AHI(C2HjO+) = 253 kcal/mole. Combining this heat of formation with the appropriate thermochemical dataz4gives a value of 11.3 ev for the ionization potential of the CH3CH20. radical. We may thus conclude (1) if CH3CH20+ were formed in a simple dissociation process involving no excess energy (which is unlikely by Stevenson’s rule,29since I(CH,CH,O) = 11.3 ev as compared with I(CH30CH2) = 7.95 e v 9 , its appearance potential would be 14.7 ev, not 11.25 as measured. (2) Again invoking Stevenson’s rule,2gthe ionization potentials of the alkoxy radicals are sufficiently high to preclude the formation of alkoxy ions in simple dissociation processes when competing dissociations forming alkoxymethyl ions ROCH2+, (RO)’CH+, etc., are available. (3) The agreement of A H f ( C H 3 0 C H 2 ~calculated ) from the appearance potentials confirms the common structure of the ion. We therefore conclude AHf(CH30CH2+) = 170 + 3 kcal/mole. Moreover, if we combine the value I(CH30CH2) = 7.9530with our result for AHr(CH3OCH2+),we arrive at the free-radical heat of formation, AHf(CH30CH2)= -14 3 kcal/mole. This figure leads to the dissociation energy D[CH30CH2-H] = 82 k 3 kcal/mole. m / e 75. (CH30)2CH+. Applying the same criteria as above (namely the unlikelihood of forming alkoxy (25) F. D. Rossini, et aL, “Selected Values of Chemical Thermodynamic Properties of the National Bureau of Standards, American Petroleum Institute, Project 44, Carnegie Institute of Technology, Pittsburgh, Pa. (26) M. S . Kharasch, Bur. Standards J . Research, 2, 359 (1929). (27) Calculated by the bond energy scheme of J . D. Cox, Telrahedron, 18, 1337 (1962). (28) I. Kanomata, Bull. Chem. SOC.Japan, 34, 1596 (1961). (29) D. P. Stevenson, Truns. Furaduy Soc., 10, 35 (1951). (30) F. P. Losing, private communication, 1964.
1355
ions in simple dissociative ionization processes when others are available) suggests strongly that m/e 75 has the structure shown in the case of formation from the molecules (CH30)zCHCH3 and (CH30)3CH. This conjecture receives further support in the agreement of the AHf values calculated in these two cases. On the other hand, in the case of m/e 75 from (CH30)2CHz, it seems at first sight that the ion could be CH3OCH20CH2+as well as (CH30)2CH+. The latter is in our opinion to be favored, however, since it is clear from the trends in Table I that the appearance potential (AP) for (CH30)2CH+should be substantially less than that for CH30CH2+(from any of the neutrals), whereas the AP for CH30CH20CH2+from ( C H 3 0 ) 2 C Hwould ~ not be expected to differ much from that of CH30CH2+ from CH30CH3. Although the value of AHf calculated for the ion of m/e 75 (Table I) from its appearance potential from (CH30)2CHz is not in precise agreement with that calculated from the other two compounds, it is likely the same ion and the figures are not outside of the combined errors. We hence assign the value AHf[(CH30)2CH+]= 109 =k 3 kcall mole. mje 105. (CH30)3C+. There is little doubt that this ion has the tertiary structure shown above when it is formed from (CH30)3CCH3and (CH30)4C. In the case of formation from (CH30)3CH, we d o not know that it is not (CH30)2CHOCH2+. The arguments by analogy used previously would again tend more toward favoring the tertiary ion structure ( D[(CH3)3C-H] is 5 kcal/mole less than D[(CH3)2CH-H]24and I[(CH,),CI3’ is 0.48 ev less than I[(CH3)nCH]32). The argument is supported by the fact that the three calculated values of AHf[(CH30)3C+]of Table I are equal within experimental error. We conclude that AHf[(CH30)3C+]= 69 f 3 kcal/mole. The other ions shown in Table I all involve loss of a C H 3 0 group and hence there is no ambiguity of their identities. Halogen-Substituted Methanes. The results of our appearance potential studies on halogen-substituted methanes are shown in Table 11. Since the dissociaTable 11. Electron-Impact Energetics in Halogen-Substituted Methanes
Re1 intensity of Molecule 50 ev
Ion ~~
CHaF+ CHFzCF3+ CH*CI+ CHCl?+ CCl,+ CH2Br+ CHd+
~
CHaF CH2F2 CHF3 CHsCI CH2Clz CHCI, CH3Br CH3I
AP, ev“
Afir(ion), kcal/mole
~~
89.5 95.6 100.0 11.4 2.0 1.3 11.1 14.1
13.25 It 0.06 13.14 i 0.02 14.03 0.06 12.98 rt 0.07 12.12 It 0.05 (?) 11.70 f 0.09(?) 12.12 + 0.09 (?) 12.08 i 0.09
197 144 106 226 205
193 217 221
The indicated uncertainties represent average deviations from the mean of replicate determinations.
species, the following thermochemical data in kcal/mole were used: AHr(CH3F) = -56.0,33 AHr(CH2F2) = - 107.2,33 AHr(CHF3) = - 1 6 5 . y 3 AHr(CH3Cl) = -20.6,33 AHr(CHzC12) = -22.4,33 AHr(CHC13) = -25.0,33 AHf(CH3Br) = and AHr(CH31) = - 5.34 CF3+. The formation of CF3+ from fluorine derivatives of methane has been studied by many workers, and the over-all result has been a set of low values (96-101 kcal/mole) and a set of high vaIues (118-121 kcal/mole) for AHf(CF3+). Our value of 106 kcal/ mole tends to support the lower set. Previously, the high value for AHr(CF3+),namely 120 kcal/mole, was to be preferred since, when it was combined with the directly measured I(CF3) = 10.1 ev,4 it gave AHf(CF3) = - 114; this value is in good agreement with the value of - I 15 f 2 determined from kinetic data.35 However, very recently Fisher, Homer, and Lossing, 3 6 from a study of the ionization potentials of larger perfluoroalkyl radicals, have concluded that the previous directly measured value5 includes >0.6 ev of excitation energy so that the true value should be -
tion process involves the loss of a hydrogen atom in all cases, there is no uncertainty in the identity of the ions. In calculating the heats of formation of the ionic (31) F. P. Lossing and J. B. deSousa, J . Am. Chem. Soc., 81, 281 (1959). (32) J. B. Farmer and F. P. Lossing, Can.J. Chem., 33, 861 (1955).
Martin, Lampe, Tuft J Methoxy- and Halogen-Substituted Methanes
1356
and which average to AHf(CC13+) = 215 =t 4 kcal/ mole. Our value of AH, = 194 kcal/mole is much too low and is undoubtedly due to uncertainties arising from the very low abundance of this ion (1.3 %) in the mass spectrum of chloroform. The low value was probably obtained because of a significant contribution from the ion CHCI3+which appears at 11.42 ev.40 Excluding our value from the other determinations and combining AHr(CC13+) with the directly measured I,(CCl3) = 8.78 ev, we arrive at AHdCCl,) = 13 f 4 kcal/mole. CHC12*. Our value of 12.12 ev for the appearance potential of this ion from CH2C12is considerably lower than the value of 13.00 ev found by Gutbier6 and Reed and Snedden.7b Our value leads to AHLCHC12+) = 205, while the higher appearance potential yields AHf(CHC12+) = 227 kcal/mole. Both values, however, are very uncertain because of the low intensity of the ion in the mass spectrum (-2 %). More meaningful heats of formation have been obtained in studies of CHCl2+ formed from CHC13’O~’’ and CHC12Br.’0 The results of these workers yield a “best” value AH,(CHCh+) = 215 f I kcal/mole, which combined with I(CHClr) = 9.30 ev leads to AHdCHCl2) = 1 A 2 kcal/mole. CH2CI+. Our appearance potential of 12.98 ev for the formation of this ion from CH3CI agrees within experimental error with the value of 12.88 found by Harrison and Shannon.’O These lead to ionic heats of formation of 226 and 224 kcal/mole, respectively, and differ from the value of 235 kcal/mole found by GutbierF by an amount somewhat outside the combined experimental error. Measurements of the appearance potential of the ion from CH2C12have been reported by G ~ i t b i e r , Harrison ~ and Shannon,’O and Reed and S n e d d e r ~ ~which ~ yield, respectively, AHf(CH2CI+) values of 235, 231, and 247. The last value is in serious disagreement with the other values reported although the reason is not clear. At any event it has been discarded in attempting to arrive at a “best” value for AHf(CH2Clt). Harrison and Shannon’O have also measured the appearance potential of CH2Cl+ from CH2C1Br and find from their result that AHf(CH2Cl+) = 229 kcal/mole. A measurement of the appearance potential of the ion from CH3CH2CI1Oyields 248 kcal/ mole for AHf(CH2ClT). This high value is expected according to Stevenson’s rule,2g since CzHj has a lower ionization potential by -0.6 ev than CH2C1, and, accordingly, the process involving CH3CI+ formation should involve excess energy. From the results discussed we suggest a “best” value of AHf(CH2CIT) = 230 i 4 kcal/mole. Combination with I(CH2C1) = 9.34 ev39yields AHi(CH2Cl) = 16 =t 4 kcal/mole. CH2J3r+. Our appearance potential from CH3Br gives AHf(CHrBr+) = 217 kcal/niole, which is unexpectedly low. This value is rejected in favor of Reed and Snedden’~’~ value of 225 determined from CH2Brz, and Gutbier’sc values of 237 from CH3Br and 238 from CH2Br?. The “best” value of AHf(CH2Br+) = 233 =t 6 kcal/mole. AHf(CH2Br) is determined by combination with I(CHsBr) = 9.3039to be 18 f 5 kcal/mole. This value is equivalent to D(CH2Br-H) = 83 A 5 kcal/mole, which appears to be satisfactorily in line with the values for D(CH2F) and D(CH2C1) shown in (40) I
